Ahead of this article, an explanation of what ‘xG’ is courtesy of the Opta website – Expected goals (xG) measures the quality of a shot based on several variables such as assist type, shot angle and distance from goal, whether it was a headed shot and whether it was defined as a big chance. Adding up a player or team’s expected goals can give us an indication of how many goals a player or team should have scored on average, given the shots they have taken.
4-4-2, 4-4-2…. I can still hear that chant some nights. It was back under the watchful eye of Peter Jackson that Accrington took us apart at home in the first half of a game, us playing a 4-3-3 with David Graham up front.
The whole of the Stacey West chanted 4-4-2 at Jackson and he changed it at half time; we went on to win 5-1. It was the ultimate proof that fans knew better and ever since I’ve been an advocate of 4-4-2. Well, I say ever since, I mean until we started winning things in a 4-2-3-1. Then I switched.
It seems now that we want 4-4-2 again, coming away from that bank of three behind Tyler Walker and giving him some help up top. it’s an interesting notion that having an extra striker means you’ll create more chances, but is it entirely true? I wanted to find out and using Wyscout’s extensive stats filters, I’ve been able to do so.
What I can do is break down our data by home and away, as well as by the formation we were playing at the time. It’ll give us a bit of an understanding as to whether or not the change that most of the supporters are behind will actually work. Yes, I know its more stats and that’s not what some of you like, but surely it’ll paint a picture of what we’re doing in certain setups?
It’s easy to dismiss stats as ‘lies’ and all that. Some will point to Coventry and Tranmere the other week in which all the numbers pointed to a Coventry win and yet Tranmere snatched a 1-0 victory. Granted, football throws up these little nuances, but on the whole, if you’re creating more chances, you have more chances to score. Eventually, if you create chances you will score, even if it means new players coming in.
What you have to bear in mind is that Wyscout is so in-depth that they monitor what you’ve moved to during a game. The only time we’ve been 4-4-2 at home this season was for the second half against Oxford, a game which skews the data a little. Away from home, we’ve used it against Pompey for 90 minutes, Rochdale and Wycombe for the second halves and from 63 minutes of our cup tie with Doncaster, which is 14% of our game time this season. The argument he probably lies in the Rochdale game; we were the better side in the second half and should have won that game. We were the better side against Pompey too and we created chances, but didn’t look likely to score any.
Whilst playing 4-4-2, as an average, our xG (expected goals) is 0.84 and our actual goals per game is 0.4. What that’s suggesting is we’re actually wasteful in 4-4-2, scoring half as many goals as we should. We average 7 shots per game whilst in this setup with an accuracy of 37%, so 7 shots with around 3 on target.
In a 4-4-1-1 we’re no less wasteful. We’ve played this in 22% of our fixtures, most successfully perhaps against Huddersfield and Everton. We did play it against Peterborough for the whole game, Blackpool for the whole game and the first half against Wycombe.
I’ve taken Everton and Huddersfield out of the numbers as the cup games against higher opposition don’t really give an indication of what we can do. I think EFL Cup fixtures are probably a bit fairer as they’ve like for like in terms of the teams we’re facing. Anyway, in the league, 4-4-1-1 delivers an xG of 0.61 and an actual goals return of 0.33; woeful numbers again. Oddly, they improve when you factor in the cup games, making this a formation that’s more suited to containing teams and hitting them on the break.
We’ve played 4-2-3-1 for the rest of the time, so I’ve broken that down by home and away. At home, we actually score more than our xG suggests we should; we bag 1.43 goals per game against an expected return of 1.11. We’re also tighter in this formation, conceding 0.43 against an xG of 0.7. Indeed, we only conceded two against Oxford in this setup.
Away from home, it’s pretty much the same story; our xG is 0.95 whilst our actual return is 1. These numbers are not spectacular, but again we do better than we should. However, the only real success we’ve had was against Rotherham, otherwise we lost the first half to Rochdale in this set-up and lost at MK Dons.
I know this is all a bit number heavy, but it does suggest that we create more, and are more clinical, when we have a bank of three behind Tyler Walker. The talk of him ‘looking isolated’ is justified, but there is a difference between looking isolated and being isolated. What I have noticed is when we feed Tyler properly, he scores. Look at the two goals against Sunderland; both came as a result of a gamble which paid off.
That’s why I’m still not throwing my weight behind 4-4-2. I don’t think we can risk playing our only two strikers at once and I fear for our style when I hear about us having a ‘target man’ to aim long for. The early matches of this season showed we’re a good footballing side and it would be criminal to miss out a player like Joe Morrell with a long ball when he’s got so much quality.
Also, 4-4-2 works best when the two strikers complement each other. Look at the best partnerships in the business; Beardsley and Lineker for England. Beardsley and Cole for Newcastle. Henry and Bergkamp for Arsenal. Suarez and Sturridge, Rooney and Van Nistelrooy even Shearer and Ferdinand. You know why they worked? Because one was a traditional ‘9’, the other a ’10’. Very few partnerships work when both are a classic ‘9’, Shearer and Sutton being one of the few. Even Emilio Butragueno and Hugo Sanchez were a 9 and 10 for Real Madrid.
John Akinde and Tyler Walker are both nines. If we wanted to play 4-4-2 it would, in my opinion, have to be Walker with Andrade or Payne.
I know you can make statistics say what you want but I genuinely went into this with an open mind. Of course, the argument can be countered. We’ve gone 4-4-2 when chasing games for instance, but shouldn’t that give us more chances? Perhaps it’s also fair to say we played much of our 4-2-3-1 whilst the former manager was in charge and the side was settled, again a fair argument.
What I would say is this; our xG is bang on our actual goals across the whole season, both are 1,13 per game. Or xG against is 1.11 but we conceded 1.33 although that number is skewed a little by that awful Oxford result. We are doing better in chance creation than the likes of Bolton (xG of 0.6) and Southend (0.94), but not as well as some of the sides we want to be tucked in with such as Gillingham (1.45) and Burton (1.82).
Pikcing the team isn’t the fan’s job, but it is something we all do with gusto. I’ve picked my starting XI for the Bolton game on the next page and I’m also going to find out which formation you would like to see, based on what you’ve read here and what you have seen for yourself.